Automatic Collision-Free Trajectory Generation for Collaborative Robotic Car-Painting

Abstract

This paper investigates the problem of collaborative robotic car-painting using a team of industrial manipulators that can be heterogeneous. Given the CAD model of the car, a collection of heterogeneous articulated robotic arms, and their corresponding fixed base positions on the factory floor/ceiling, the objective is to generate a collection of joint trajectories for each robot in a computationally efficient manner such that the car body can be painted by the nozzles attached to the arms while collisions during the painting process are avoided. Our solution to this computationally intensive collaborative coverage path planning relies on decoupling the collision avoidance from the coverage path planning by exploiting the inherent two-dimensional structure of the problem. In particular, our algorithm relies on partitioning the reachable space of the forearms of these robots, projecting the resulting volumes of intersection on the sides and the top of the car body, and performing the coverage planning on the resulting projected volumes. Simulation results on several industrial arms that are collaboratively painting a Ford Motor Company F-150 truck demonstrate the effectiveness of our proposed solution.